CHILDS GLACIER 1910, when it was moving at the rate of 80 to 40 feet a day, or more. If the glacier terminus had not been swept by the Copper River, the rapidly-moving middle of the glacier certainly would have advanced this distance. The slowly-moving northern margin did advance 1869 feet, at a rate which increased from practically zero in 1909 till it reached £ to 8 feet a day in 1910. Advance here was possible because this part of the glacier ends on the land where the river could not cut it back. At the end of the summer of 1910 this northern portion of the glacier was only 1571 feet from the railway bridge. If the river had not had its normal summer rise, or had the advance occurred in the winter when the water was low and the river weak, the middle of the glacier would surely have advanced a good part, if not all, of the distance to the railway. As it was the ice front did move forward more than 600 feet before the rise came. It is evident, therefore, that the preservation of the railway and bridge during the summer of 1910 was due to the river, the very thing which necessitates the bridge. The relationships of the variations of the glacier to stages of river level are as follows: OSCILLATIONS OF MIDDLE OF CHTLDB GLACIER AND OF COPPER RIVER LEVEL IN 1910 Date Variations of Glacier Stage of River 1909 to JuneS, 1910 June 3 to Aug. 11 Aug. 11 to Aug. 17 Aug. 17 to unknown date Unknown date to Oct. 5,1910 Advance 275-650 feet Retreat 415 feet Retreat 160 feet Retreat, continued Advance 325 feet 2 Fall l 12 feet, May 6 to June 3 Rise 6 feet Level about stationary Rise IT^ feet Fall 9 feet Resumption of Normal Conditions in 1911. Upon the return of the junior author to Childs Glacier in June, 1911, it was found that the slowing down of the advance of the previous year (see table, p. 406) had been continued. The northern margin of the glacier had advanced only 97 feet toward the railway bridge (PI. CLIIE), and the rate of motion had been reduced from 44-100 of a foot a day in September-October, 1910, to 38-100 of a foot a day in June, 1911. Conditions were not yet quite normal, however, for under usual conditions this portion of the glacier terminus has little motion, and the amount of ice brought forward is so nearly balanced by melting, that the ice edge either remains stationary or retreats. In 1911, however, the northern margin was still advancing slowly, but appreciably, into the forest, a tape line measurement proving an advance of approximately 1 foot in the 3 days between June 16 and 19. This portion of the ice front was almost as severely crevassed as in the previous summer. There was much less tormina.! moraine and in most places there was none, though at one point the moraine was 5 or 6 feet high. On the river bank there was a low, partly-submarginal moraine, made up of rounded bowlders and including some dead alders. The stream draining the marginal lake flowed through the forest parallel to the ice front, having abandoned its subglacial channel since the previous autumn. That the rate of motion decreased also in mid-glacier is made plain by the fact that 1 Low water throughout most of fall and winter, with rise from March 18 to May 6. i Plus unmeasured retreat after August 17.